StyleAdaptedLM: Enhancing Instruction Following Models with Efficient Stylistic Transfer

Apoorv Saxena; Debraj Basu; Meghanath M Y; Pritika Ramu; Varsha Sankar

REVIEW 5 major objections 6 minor 33 references

StyleAdaptedLM: Enhancing Instruction Following Models with Efficient Stylistic Transfer

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that a LoRA adapter trained on unannotated text in one style can be merged into a separate instruction-tuned model to produce style-consistent, instruction-adherent text without paired data.

desk verdict A practical LoRA-merge recipe for brand style that works in pilot experiments, but missing controls and a test-set-tuned ratio keep the headline claim from being fully established. read the letter →

arxiv 2507.18294 v1 pith:EXG3R3BW submitted 2025-07-24 cs.CL

Pritika Ramu , Apoorv Saxena , Meghanath M Y , Varsha Sankar , Debraj Basu This is my paper · ORCID

classification cs.CL

keywords stylistictransferlow-rankadaptation(LoRA)instructionfollowingmodelmergingunstructuredtextbrandvoiceenterprisecommunication

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

StyleAdaptedLM claims that a low-rank adapter trained on plain, unannotated text in a target style—a person's emails, a news agency's articles, or an enterprise's marketing mail—can be added directly to a separate instruction-tuned model, giving that model the style while leaving its instruction-following mostly intact. In tests across three model families and several style corpora, strict instruction-following accuracy stays within roughly five to nine percentage points of the untouched instruct model, and content fidelity is best or near-best among the compared methods. A human evaluation by an experienced copywriter rates the merged outputs higher for ease of publication and overall quality than few-shot prompting. The authors present this as a practical way for enterprises to manage many brand voices as small adapters without paired instruction-response data or full retraining.

What carries the argument

The central object is the LoRA adapter and its additive merge identity. LoRA (low-rank adaptation) is a parameter-efficient fine-tuning method that learns only a low-rank weight update $\Delta W = AB$, with $A \in \mathbb{R}^{d_{\text{out}} \times r}$, $B \in \mathbb{R}^{r \times d_{\text{in}}}$, and $r \ll d_{\text{in}}$, instead of updating full model weights. StyleAdaptedLM trains this update on a base model with next-token completion over style-annotated text, then forms $W_{\text{merged}} = W_{\text{instruct}} + \Delta W$ by adding the update to a separate instruction-tuned model. The paper's argument is that because the update is low-rank and small, the instruction model's parameter-space redundancy absorbs the shift without losing its abilities; this additivity is the machinery that carries the claimed style transfer.

What would settle it

Merging one style adapter into an instruction-tuned model from a different model family than the base it was trained on (for example, training the adapter on a Llama base and merging into Qwen-Instruct) and measuring IFEval and style-classifier scores would settle whether the LoRA delta is a portable style shift; a large IFEval drop or incoherent outputs would falsify the claim.

Watch

Extended reading notes

Core claim

The paper's discovery is that style can behave as a modular addition in the weight space of a language model. LoRA adapters are trained on a base pretrained model using a next-token objective over corpus text prefixed with a style tag such as 'News article written by [[BBC]].' The trained adapter update $\Delta W$ is then added to the weights of a different instruction-tuned model: $W_{\text{merged}} = W_{\text{instruct}} + \Delta W$. The resulting model writes in the target style—BBC versus CNN spelling and tone, an individual Enron employee's email register, an enterprise's marketing conventions—while keeping instruction-following nearly intact: IFEval strict accuracy drops by about 5–9 points, and tinyMMLU reasoning scores do not change. The paper reports that this merge outperforms direct LoRA fine-tuning of the instruct model and matches or slightly exceeds model soups, while producing the best or near-best ROUGE-1 content scores.

Load-bearing premise

The load-bearing premise is that a small style adjustment learned on one pretrained model can be glued onto a different, instruction-trained version of the same model and will change writing style without degrading instruction-following, an additivity claim the paper supports with intuition rather than proof.

Editorial extensions

If this is right

Enterprises can maintain many brand voices as small, swappable LoRA adapters trained on past marketing emails and merge whichever one is needed at serving time, instead of hosting a full model per brand.
Styles can be learned from corpora that lack instruction-response formatting, removing the paired-data bottleneck that blocks direct style fine-tuning of instruction models.
The method keeps the instruction model's weights frozen, making it cheaper than model soups or full fine-tuning while matching or slightly beating those baselines on instruction adherence.
Explicit style annotations during adapter training measurably improve style adherence compared with training on the same corpus without annotations.
Avoiding few-shot prompting sidesteps the measured decline in instruction-following that longer prompts cause, while still capturing style from examples.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural untested extension is to view each adapter delta as a vector in parameter space and interpolate or add several style adapters to produce blended voices; the paper only demonstrates single-style merges.
If the additivity claim generalizes, the same pipeline could transfer conventions beyond prose style—such as code-comment tone, legal boilerplate, or product documentation—wherever an unstructured corpus exists but paired instruction data does not; the paper does not test these domains.
The paper's own limitation note implies a testable risk: if the style corpus contains factual claims or biases, the merged model may adopt them as part of the 'style,' so a stress test with a corpus of planted false statements would separate stylistic uptake from factual contamination.
The annotation ablation suggests that richer style labels (for example, nested brand, format, and tone tags) could act as a control knob for transfer strength, though the paper compares only annotated versus unannotated training.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: StyleAdaptedLM is a sensible, low-cost recipe for transferring style to instruction-following models — train a LoRA adapter on a base model with plain style-annotated text, then merge the delta into the instruct model. The evaluation is broad (three model families, four corpora) and the results are directionally supportive: IFEval stays within 5-9% of the original instruct model, style F1 is near the few-shot prompt baseline, and ROUGE-1 improves. That is worth a serious look for anyone doing enterprise personalization or PEFT.

What's genuinely new is the specific combination. Training a LoRA on unstructured text and adding it to a different model has been explored piecewise (LoRA composition, model soups, Liu et al.'s style adapters), but I don't know of prior work that tests this exact cross-model transplant across news, email, Reddit, and marketing data. The modularity angle — one base adapter per brand, merged into any instruct model — is practically appealing.

The soft spots are real, though. The stress-test note is on target: the paper never shows that the instruct model given the same concrete style label (e.g., 'Write a [[BBC]] article') without the adapter fails to produce style. The Table 6 baseline uses the generic placeholder '[[news source]]', which doesn't control for the cue. The merge ratio (2:1) is chosen by testing on the evaluation sets, so the reported IFEval drop is an optimistic estimate. There are no error bars despite 'averaging over three runs,' and the human eval is one expert rater. These don't sink the method, but they do prevent the abstract's 'without sacrificing task performance' from being established with confidence. The paper honestly flags the pretraining-overlap concern in §7, which is good, but it doesn't resolve it.

I'd send this to reviewers. It's a legitimate applied contribution with a clean enough design that the missing controls can be fixed with a modest revision — add a control with the exact label, choose the ratio on a held-out split, report variance, and release code/data. The mechanism in §3.4 is hand-wavy, but the empirical results are what matter.

For a reading group: maybe. It's not deep, but it's a useful case study in how evaluation confounds can obscure an otherwise plausible PEFT trick.

Recommendation: send to peer review, expect revision.

Referee Report

5 major / 6 minor

Summary. StyleAdaptedLM trains LoRA adapters on a base model using style-annotated unstructured text and then adds the learned weight delta to a separate instruction-tuned model. The paper reports experiments across Llama-3.1-8B, Mistral-7B, and Qwen2.5-7B on Enron, Reddit, CNN/BBC, and a proprietary marketing email corpus, measuring instruction following with IFEval and tinyMMLU, style with an authorship-attribution classifier and one professional rater, and content fidelity with ROUGE-1. The headline claim is that the method yields style-consistent, instruction-adherent outputs without paired instruction-response data and with only a small IFEval drop relative to the unmodified instruct model.

Significance. If the central claim is correct, the method is practically useful: one style adapter trained on raw text can be modularly attached to an existing instruct model, avoiding separate fine-tunes per brand. Strengths include the breadth of model families and datasets, the comparison against direct LoRA fine-tuning and model soups, the style-annotation ablation, the IFEval benchmark, and qualitative examples showing region-specific spelling and template conventions. The main gaps are that the robustness claim is not tested on styles absent from pretraining, that the merge ratio is chosen on the test set, and that the mechanism of cross-model LoRA transfer is asserted rather than demonstrated; these gaps are fixable but currently limit the strength of the conclusions.

major comments (5)

[Section 7] Section 7 explicitly concedes that the evaluation corpora may overlap with pretraining data. Because BBC, CNN, Enron, and Reddit are public and very likely seen by Llama, Mistral, or Qwen during pretraining, the abstract's 'robust stylistic customization' is not yet established. Please add at least one evaluation on a style that cannot be in pretraining (e.g., a newly collected proprietary corpus or a synthetic style constructed after the knowledge cutoff) and report the same IFEval and attribution metrics; if that is impossible, revise the robustness claim accordingly.
[Section 4.2] Section 4.2 reports that the 2:1 merge ratio was selected because it 'performed best' among 1:2, 1:1, and 2:1, but no separate validation split is used. Since the ratio is tuned on the same test prompts that produce the reported IFEval and attribution numbers, the reported drops in instruction-following accuracy are optimistic estimates. Use a held-out split for ratio selection and report the test numbers for the chosen ratio, or report the results for all three ratios.
[Section 3.4] Section 3.4 asserts that LoRA shifts are 'minimal' and 'absorbed' by parameter-space redundancy, but no measurement or analysis supports this mechanism. Table 2's contrast between direct LoRA fine-tuning and StyleAdaptedLM is suggestive, but the paper should provide a direct diagnostic: for example, the per-layer norm of the merged delta, a control where the same merge is applied to the base model, or an ablation replacing the style delta with a random or instruction-tuned delta. Without this, the central mechanism remains a conjecture.
[Tables 2-3 vs Table 6] Table 6 is a useful control for news styles, showing that the concrete label alone does not produce high style accuracy in the unmodified instruct model. However, the main results in Tables 2-3 do not include the analogous no-fine-tuning condition with concrete style labels for Enron senders, Reddit users, or the enterprise; the generic 'Instruct Model (No FT)' row is not sufficient to rule out prompt-cue effects in those styles. Please add the same concrete-label control for every style.
[Section 4.3.2, Table 3] The authorship-attribution classifier in Section 4.3.2 may be exploiting topical content rather than style; the paper acknowledges this, but since Table 3 is the primary style-adherence evidence, the confound is load-bearing. Please include a content-controlled evaluation where the same note is rendered in multiple target styles so that the classifier must rely on style, or report per-topic classifier scores.

minor comments (6)

[Section 4.2] No error bars or significance tests are reported despite 'averaging results over three runs'; please report standard deviations or confidence intervals for the main tables.
[Section 4.3.2, Figure 2] The human evaluation relies on a single rater; this should be stated as a limitation and, ideally, supplemented with multiple raters or inter-annotator agreement.
[Appendix D] Appendix D's table header says 'BrandAdaptedLM' while the method is called StyleAdaptedLM elsewhere; please correct the inconsistency.
[References] The duplicate Brown et al. 2020a and 2020b references should be consolidated, and incomplete access dates (e.g., OpenAI 2023) should be filled in.
[Tables] Table 2's layout is garbled in the submitted PDF, with row labels and numbers joined; please regenerate the tables so that each method row and model column is clearly aligned.
[Section 4.1] The marketing email corpus is anonymized and not released, so the enterprise results cannot be reproduced; consider releasing a redacted version or a synthetic substitute, and describe the GPT-4o prompt-generation procedure for the validation sets.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the LoRA-transplant claim is empirical and evaluated on external benchmarks; the confounds noted in the paper are validity concerns, not derivation-circularity.

full rationale

StyleAdaptedLM's central claim is that a LoRA adapter trained on unstructured style corpora with a base model can be merged into a separate instruction-tuned model to impart style while preserving instruction following. The training objective is next-token prediction on the style-annotated corpus, and the merged weights are defined as W_merged = W_instruct + Delta_W(style|base). Neither the adapter weights nor the merge rule are constructed from IFEval scores, ROUGE, or the authorship classifier; those are external measurements applied after training. The authorship-attribution classifier is trained on the target corpora and applied to generated text, which raises a task-validity question about whether classifier agreement reflects style or content, and the paper explicitly acknowledges this difficulty while adding ROUGE-1 as a content check. That is an evaluation concern, not a circular derivation. The 2:1 merge ratio is selected empirically; if it was selected on the same test sets, the reported numbers would be optimistic, but the reported IFEval, style, ROUGE, and human results are measurements, not quantities forced by the ratio through an equation. The Section 3.4 compositionality rationale ('small shifts are easily absorbed due to the redundancy in the parameter space') is an unproven assumption rather than a circular step. The paper's own Section 7 limitation flags potential dataset overlap with pretraining data and defers evaluation on novel or proprietary corpora, which is an external-validity confound rather than a definitional reduction. There are no load-bearing self-citations or imported uniqueness theorems. Hence no circularity is exhibited; score 0.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The central claim rests on a small number of hand-set parameters, most importantly the merge ratio and undisclosed LoRA hyperparameters, plus a geometric assumption that base-model LoRA deltas transfer cleanly to instruct models. No new entities are introduced. The method is largely a combination of established components, so the contributory part is the empirical configuration rather than a new theoretical object.

free parameters (3)

merge ratio = 2:1 (style-adapted base : instruct)
Chosen because it 'performed best' in experiments; no ablation of other ratios is shown, so the choice may be tuned to the evaluation sets.
style annotation prefixes = e.g., 'News article written by [[CNN]].'
Hand-designed prompt prefixes; the ablation shows annotations help, but the specific wording is a design choice that could affect results.
LoRA hyperparameters = not reported
Rank, learning rate, epochs, and batch size are not stated, so the training recipe is a hidden set of free parameters needed for replication.

assumptions (3)

ad hoc to paper LoRA update matrices trained on a base model can be added to a different instruct model without disrupting its capabilities.
Section 3.4 justifies this with a heuristic about low-rank shifts being small and absorbed by parameter redundancy; no proof or systematic analysis is given.
domain assumption UAR embeddings and an SVM classifier provide a valid measure of style adherence.
Section 4.3.2 relies on prior work for authorship attribution without validating that the classifier distinguishes style from content; ROUGE scores are used as a partial check.
domain assumption GPT-4o generated neutral paraphrases and validation prompts are sufficient and unbiased proxies for instruction-based generation.
Section 4.1 uses GPT-4o to create inputs and neutral versions of Reddit comments, but the quality and neutrality of these generations are not independently verified.

how reviews work

0 comments

Cite this review

Pith. "Pith review of StyleAdaptedLM: Enhancing Instruction Following Models with Efficient Stylistic Transfer." pith.science (2026). https://pith.science/paper/EXG3R3BW

@misc{pith2026250718294,
  author       = {Pith},
  title        = {Pith review of: StyleAdaptedLM: Enhancing Instruction Following Models with Efficient Stylistic Transfer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EXG3R3BW}},
  note         = {Machine review of arXiv:2507.18294}
}

read the original abstract

Adapting LLMs to specific stylistic characteristics, like brand voice or authorial tones, is crucial for enterprise communication but challenging to achieve from corpora which lacks instruction-response formatting without compromising instruction adherence. We introduce StyleAdaptedLM, a framework that efficiently transfers stylistic traits to instruction-following models using Low-Rank Adaptation (LoRA). LoRA adapters are first trained on a base model with diverse unstructured stylistic corpora, then merged with a separate instruction-following model. This enables robust stylistic customization without paired data or sacrificing task performance. Experiments across multiple datasets and models demonstrate improved stylistic consistency while preserving instruction adherence, with human evaluations confirming brand-specific convention uptake. StyleAdaptedLM offers an efficient path for stylistic personalization in LLMs.

Figures

Figures reproduced from arXiv: 2507.18294 by the authors.

**Figure 1.** StyleAdaptedLM Framework. Inputs: user instruction and corpus (e.g., brand’s past marketing emails). Method: A LoRA adapter, pre-trained on the corpus with a base LLM, is merged with an instructiontuned LLM. Output: Instruction-adherent, style-aligned text, potentially including corpus-learned characteristics. in enabling LLMs to adopt these diverse stylistic identities without degrading their fundamental instruc… view at source ↗

**Figure 2.** Human ratings (Likert scale 1–5) assessing [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 10 canonical work pages

[1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[3]

Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M Ziegler, Jeffrey Wu, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Sc...

work page 2020
[4]

Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gr...

2020
[5]

Corinna Cortes and Vladimir Vapnik. 1995. Support-vector networks. Machine learning, 20(3):273--297

1995
[6]

Derek Greene and P \'a draig Cunningham. 2006. Practical solutions to the problem of diagonal dominance in kernel document clustering. In Proceedings of the 23rd international conference on Machine learning, pages 377--384

work page 2006
[7]

Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021. Measuring massive multitask language understanding. Proceedings of the International Conference on Learning Representations (ICLR)

2021
[8]

Karl Moritz Hermann, Tomáš Kočiský, Edward Grefenstette, Lasse Espeholt, Will Kay, Mustafa Suleyman, and Phil Blunsom. 2015. Teaching machines to read and comprehend. Advances in neural information processing systems, 28

work page 2015

Show all 33 references

[9]

Zachary Horvitz, Ajay Patel, Kanishk Singh, Chris Callison-Burch, Kathleen McKeown, and Zhou Yu. 2024. https://doi.org/10.18653/v1/2024.findings-emnlp.781 T iny S tyler: Efficient few-shot text style transfer with authorship embeddings . In Findings of the Association for Comp...

2024 doi
[10]

Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2022. https://openreview.net/forum?id=nZeVKeeFYf9 Lo RA : Low-rank adaptation of large language models . In International Conference on Learning Representations

2022
[11]

Vetrov, and Andrew Gordon Wilson

Pavel Izmailov, Dmitrii Podoprikhin, Timur Garipov, Dmitry P. Vetrov, and Andrew Gordon Wilson. 2018. http://dblp.uni-trier.de/db/conf/uai/uai2018.html#IzmailovPGVW18 Averaging weights leads to wider optima and better generalization . In Proceedings of the Thirty-Fourth Confer...

2018
[12]

Aleem Khan, Elizabeth Fleming, Noah Schofield, Marcus Bishop, and Nicholas Andrews. 2021. https://doi.org/10.18653/v1/2021.naacl-main.415 A deep metric learning approach to account linking . In Proceedings of the 2021 Conference of the North American Chapter of the Association...

2021 doi
[13]

Bryan Klimt and Yiming Yang. 2004. The enron corpus: A new dataset for email classification research. In European Conference on Machine Learning, pages 217--226. Springer

2004
[14]

Xiang Lisa Li and Percy Liang. 2021. https://doi.org/10.18653/v1/2021.acl-long.353 Prefix-tuning: Optimizing continuous prompts for generation . In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conferen...

2021 doi
[15]

Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang

Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2024 a . https://doi.org/10.1162/tacl_a_00638 Lost in the middle: How language models use long contexts . Transactions of the Association for Computational Linguistics,...

2024 doi
[16]

Xiao Liu, Kaixuan Ji, Yicheng Fu, Weng Lam Tam, Zhengxiao Du, Zhilin Yang, and Jie Tang. 2021. P-tuning: Prompt tuning can be comparable to fine-tuning. arXiv preprint arXiv:2103.10385

2021 arXiv
[17]

Xinyue Liu, Harshita Diddee, and Daphne Ippolito. 2024 b . https://aclanthology.org/2024.inlg-main.34 Customizing large language model generation style using parameter-efficient finetuning . In Proceedings of the 17th International Natural Language Generation Conference, pages...

2024
[18]

Huan Ma, Changqing Zhang, Yatao Bian, Lemao Liu, Zhirui Zhang, Peilin Zhao, Shu Zhang, Huazhu Fu, Qinghua Hu, and Bingzhe Wu. 2023. http://arxiv.org/abs/2303.13217 Fairness-guided few-shot prompting for large language models

2023 arXiv
[19]

Vrinda Matena and Colin Raffel. 2022. https://arxiv.org/abs/2209.14865 Merging models with different initialization paths . arXiv preprint arXiv:2209.14865

2022 arXiv
[20]

Sewon Min, Mike Lewis, Hannaneh Hajishirzi, and Luke Zettlemoyer. 2022. https://arxiv.org/abs/2202.12837 Rethinking the role of demonstrations: What makes in-context learning work? In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP)

2022 arXiv
[21]

Sourabrata Mukherjee and Ondrej Dušek. 2024. http://arxiv.org/abs/2407.14822 Text style transfer: An introductory overview

2024 arXiv
[22]

OpenAI. 2023. Gpt-4. https://platform.openai.com/docs/models/gpt-4. Accessed: YYYY-MM-DD

2023
[23]

Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul F Christiano, Jan Leike, a...

2022
[24]

Felipe Maia Polo, Lucas Weber, Leshem Choshen, Yuekai Sun, Gongjun Xu, and Mikhail Yurochkin. 2024. http://arxiv.org/abs/2402.14992 tinybenchmarks: evaluating llms with fewer examples

2024 arXiv
[25]

Sudha Rao and Joel Tetreault. 2018. https://doi.org/10.18653/v1/N18-1012 Dear sir or madam, may I introduce the GYAFC dataset: Corpus, benchmarks and metrics for formality style transfer . In Proceedings of the 2018 Conference of the North A merican Chapter of the Association ...

2018 doi
[26]

Emily Reif, Daphne Ippolito, Ann Yuan, Andy Coenen, Chris Callison-Burch, and Jason Wei. 2022. https://doi.org/10.18653/v1/2022.acl-short.94 A recipe for arbitrary text style transfer with large language models . In Proceedings of the 60th Annual Meeting of the Association for...

2022 doi
[27]

Rivera-Soto, Olivia Elizabeth Miano, Juanita Ordonez, Barry Y

Rafael A. Rivera-Soto, Olivia Elizabeth Miano, Juanita Ordonez, Barry Y. Chen, Aleem Khan, Marcus Bishop, and Nicholas Andrews. 2021. https://doi.org/10.18653/v1/2021.emnlp-main.70 Learning universal authorship representations . In Proceedings of the 2021 Conference on Empiric...

2021 doi
[28]

Bakhtiyar Syed, Gaurav Verma, Balaji Vasan Srinivasan, Anandhavelu Natarajan, and Vasudeva Varma. 2020. Adapting language models for non-parallel author-stylized rewriting. In Proceedings of the 34th AAAI Conference on Artificial Intelligence

2020
[29]

Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, W...

2023 arXiv
[30]

Smith, Daniel Khashabi, and Hannaneh Hajishirzi

Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A. Smith, Daniel Khashabi, and Hannaneh Hajishirzi. 2023. https://doi.org/10.18653/v1/2023.acl-long.754 Self-instruct: Aligning language models with self-generated instructions . In Proceedings of the 61st Annual Mee...

2023 doi
[31]

Mitchell Wortsman, Gabriel Ilharco, Samir Ya Gadre, Rebecca Roelofs, Raphael Gontijo-Lopes, Ari S Morcos, Hongseok Namkoong, Ali Farhadi, Yair Carmon, Simon Kornblith, and Ludwig Schmidt. 2022. https://proceedings.mlr.press/v162/wortsman22a.html Model soups: averaging weights ...

2022
[32]

Chiyu Zhang, Honglong Cai, Yuezhang Li, Yuexin Wu, Le Hou, and Muhammad Abdul-Mageed. 2024. https://doi.org/10.18653/v1/2024.naacl-srw.21 Distilling text style transfer with self-explanation from LLM s . In Proceedings of the 2024 Conference of the North American Chapter of th...

2024 doi
[33]

Jeffrey Zhou, Tianjian Lu, Swaroop Mishra, Siddhartha Brahma, Sujoy Basu, Yi Luan, Denny Zhou, and Le Hou. 2023. Instruction-following evaluation for large language models. arXiv preprint arXiv:2311.07911

2023 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [3]

Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel M Ziegler, Jeffrey Wu, Clemens Winter, Christopher Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Sc...

work page 2020

[4] [4]

Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gr...

2020

[5] [5]

Corinna Cortes and Vladimir Vapnik. 1995. Support-vector networks. Machine learning, 20(3):273--297

1995

[6] [6]

Derek Greene and P \'a draig Cunningham. 2006. Practical solutions to the problem of diagonal dominance in kernel document clustering. In Proceedings of the 23rd international conference on Machine learning, pages 377--384

work page 2006

[7] [7]

Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021. Measuring massive multitask language understanding. Proceedings of the International Conference on Learning Representations (ICLR)

2021

[8] [8]

Karl Moritz Hermann, Tomáš Kočiský, Edward Grefenstette, Lasse Espeholt, Will Kay, Mustafa Suleyman, and Phil Blunsom. 2015. Teaching machines to read and comprehend. Advances in neural information processing systems, 28

work page 2015